Rainfall seasonality shapes belowground root trait dynamics in an Amazonian tropical rainforest: A test of the stress-dominance hypothesis
Data files
Dec 24, 2024 version files 165.32 KB
-
README.md
2.73 KB
-
Root_Morphology_FACE_2016-2017.csv
162.59 KB
Abstract
The stress-dominance hypothesis (SDH) predicts that trait variation at the community level increases with the availability of limiting resources, driving spatial and temporal patterns in aboveground plant functional trait expression. Here, we test the assumption that the SDH also applies to fine roots responding to spatial and temporal fluctuations in soil resource availability. We monitored fine root mass and functional root traits associated with resource acquisition, i.e., specific root length (SRL), specific root tip abundance (SRTA) and branching index (BI), and traits related to stress tolerance, like root diameter (RD) and tissue density (RTD) in a Central Amazonian tree community. To test for spatial differences in root traits we separated the uppermost organic (O-A horizon, 0-5 cm) and mineral soil (B horizon, 5-15 cm) layers, and for temporal fluctuations we investigated the relationship of precipitation on community-level root variation over a period of 27 months. In accordance with the SDH, we found that fine roots in the O-A horizon have on average 15% higher SRL, 23% higher BI, 32% higher SRTA, and 15% lower RTD than those in the B horizon. Similarly, precipitation shifted the community over time to higher mean SRL, BI, and SRTA (r= 0.92, 0.84 and 0.94, p<0.0001 respectively), although trait shifts occurred in the trimester after the rainy season onset, revealing a time-lag between rainfall patterns and community response. We also detected a positive increase in trait range for SRL and SRTA with lagged precipitation (r=0.90 and 0.79, p<0.0001). On the other hand, traits related to stress showed a weaker negative relationship with instantaneous precipitation (r =-0.7 and -0.57, p = 0.046 and p =0.1 for RD and RTD respectively). Our results supported the SDH predictions that root systems will become more acquisitive in areas with more resources, and that the community will shift to more acquisitive but also broader trait dispersion as hydric stress decreases. We conclude that although higher resource availability may increase competition for acquisition, trait overdispersion seems to promote species coexistence. Our results highlight how dynamic root systems can be in response to environmental cues, cautioning the common practice of making conclusions about root traits adaptations to environmental gradients based on a single sampling observation.
README: Rainfall seasonality shapes belowground root trait dynamics in an Amazonian tropical rainforest: A test of the stress-dominance hypothesis
https://doi.org/10.5061/dryad.0k6djhbb0
Description of the data and file structure
Root morphological trait measurements
Each selected fine root sample was scanned using a high-resolution flatbed scanner (600 DPI resolution, 256-level grayscale, TIFF format; Epson Scanner Perfection V700 Photo, USA) and analyzed using WinRhizo software (2007 Pro version, Instrument Regent, Quebec, Canada, details in Appendix S1) following Bouma et al. (2000). From the image analysis we extracted the total length, and total number of tips per root sample. Using the dry mass per root sample, we calculated SRL, SRTA, and BI as the trait positively associated with root acquisition function. In addition, using volume estimations from the software we estimated RTD as the trait positively associated with stress tolerance, and average root diameter from the images as RD, representing the tradeoff between mycorrhizal dependency and soil exploratory ability (Leuschner et al., 2004; Meinen et al., 2009).
Files and variables
File: Root_Morphology_FACE_2016-2017.csv
Variables
- Sampling= Consecutive order of the sampling events;
- Date=Month/Year when filed sampling event was performed;
- Core= Location where samples were taken, firts number refers to distance from the start of the transect, second numbler the replicate at that sampling point. In each point three random locations were cored as replicateds;
- WhinRhizo= Random Number assigned to the sample to be scanned from the total group of roots extracted from the core;
- Depth= Depth were the samples was obtained, categorical: 0-5 cm (O-A horizon) 5-15 cm (B horizon);
- Distance= distance of sampling point from the start of the transect;
- specific root length (SRL, m/g), average root diameter (AvgDiam, mm), and root tissue density (RTD,cm3/g) are main components of the root economic spectrum; specific root area (SRA, cm^2/g) was highly positively correlated with SRL thus not used for the analysis.
- the branch configuration of the root systems expressed as branchinness (BI, tips/cm), and specific root tip abundance (SRTA, tips/mg).
- LengPerVol makes an estimate of the total volume of soil occupied by the root, this variable was not used for analysis. Weight refers to the dry mass of the root system;
- RootVolume= Estimated volume of each scanned root sample, calculated as the summary of the volume of all links identified by the software WhinRhizo.
- Count= Total number of individual root systems scanned in each soil sample.
Methods
Site description:
The study was carried out in the Cuieiras Reserve at ZF2, ca. 60 km north of Manaus, Amazonas, Brazil (latitude S 2° 35' 40'', longitude W 60° 12' 28''). The vegetation is an old growth lowland rainforest with a 40 m tall canopy dominated by the families Fabaceae, Sapotaceae, Burseraceae, and Lecythidaceae (Higuchi et al., 1998; da Silva et al., 2011). The mean air temperature is 26°C and the mean annual precipitation is 2400 mm (Araújo 2002; Figure 2). Soils are classified as Geric Ferralsols with a pH of 3.9 on average (Quesada et al., 2010). Nutrient availability in the mineral soil is low, particularly for P, with available P varying from 7 to 16 mg kg-1 (Schaap et al., 2021). The precipitation during the study period was obtained from a meteorological station located 2 km from the study site (Tower K34, Cordeiro et al., 2020). During the studied period, precipitation followed the climate seasonality of the area (Fontes et al., 2018; Cordeiro et al., 2020; Figure 2) and increased from March to May 2016 and from January to May 2017. Dry periods occurred from September to December 2016 and more markedly from July to November 2017. For this study, accumulated precipitation was estimated as the total amount of rain recorded in the three months before the soil cores were sampled.
Soil and litterfall collection:
To measure changes in root morphology over time, soils were sampled sequentially from 18 locations along a 500 m transect between February 2016 and February 2018 every three-months (9 sampling dates). Each location was separated by a gap of approximately 30 m, with three cores collected per location, 10 m away from each other and perpendicular to the main transect, thus amounting to a total of 54 cores per sampling date. After removing loose senescent leaves from the topsoil, cores were taken with a custom-made steel soil corer (10 cm in diameter) down to 15 cm soil depth. Each sample was sealed in a plastic bag and taken to the lab. In the lab, samples were separated into two different soil layers: the organic horizon — including the surface mat of roots, litter and humus (0-5 cm) — and the mineral horizon (5-15 cm), hereafter named the O-A and B horizons respectively. Soil cores were left in water overnight and sieved in a 2 mm mesh, while entire roots were carefully separated and classified as dead or alive based on the elasticity and color of the root system (Valverde-Barrantes et al., 2013). Roots were classified as live or dead based on friability, color, and the presence of a stele following Meinen et al. (2009). Live fine roots < 1 mm in diameter were classified using an electronic caliper and pooled for biomass assessment. A subset of fine roots was randomly selected for morphological analysis. We defined our sampling unit within cores as fine root clusters ~10 cm in length, which generally included two to three orders of non-lignified roots attached to a woodier central segment, hereafter referred to as fine root sample. On average, 7 ± 2 fine root samples were extracted from each core (~20 % of the total biomass) for scanning. After scanning, all samples were oven-dried for 48 h at 65°C to constant weight (Valverde-Barrantes et al., 2013). Fine root biomass encompasses all fine roots, including those used for morphological analysis.